Learning and Individual Differences 39 (2015) 81–88

Contents lists available at ScienceDirect

Learning and Individual Differences

journal homepage: www.elsevier.com/locate/lindif

Are balanced groups better? Belbin roles in collaborative

learning groups☆
Nicoleta Meslec a,⁎, Petru Lucian Curşeu b,c,1
a
Department of Organisation Studies, Tilburg University, PO Box 90153, 5000 LE Tilburg, The Netherlands
b
Department of Psychology, “Babeş-Bolyai” University, Cluj-Napoca, Romania
c
Department of Organisation Studies, Tilburg University, PO Box 90153, 5000 LE Tilburg, The Netherlands

a r t i c l e i n f o a b s t r a c t

Article history: In a sample of 459 students organized in 84 groups this study tests the impact of group role balance on teamwork
Received 6 March 2013 quality and three performance indicators in collaborative learning groups (group cognitive complexity, perceived
Received in revised form 26 May 2014 performance and objective performance). The results show that group role balance positively predicts group
Accepted 29 March 2015
performance in preliminary phases of the group project but not in later phases. Moreover, group role balance
positively predicts group cognitive complexity and is negatively related to teamwork quality. The results hold
Keywords:
Group role balance
only when role balance is conceptualized as a configural property of groups instead of a sum of individual
Performance roles. The findings of the study have implications for the design of collaborative learning groups.
Collaborative learning groups © 2015 Published by Elsevier Inc.
Group cognitive complexity

1. Introduction adopt specific patterns of behavior in interpersonal interactions and

Modern organizations use groups to perform a variety of complex
tasks (Tannenbaum, Mathieu, Salas, & Cohen, 2012), therefore, next to
job-related knowledge and expertise, teamwork skills become important
elements in personnel selection across a variety of organizational fields
(Burch & Anderson, 2004; O'Neil, Allerd, & Baker, 1997; Stevens &
Campion, 1999; Zedeck & Goldstein, 2000). As a consequence, education-
al programs extensively use collaborative learning to help students devel-
op teamwork skills (Curşeu, Janssen, & Raab, 2012) and acquire specific
curricular knowledge (Curşeu, 2011; Haugwitz, Nesbit, & Sandmann,
2010; McCune & Entwistle, 2011). It becomes therefore important to
identify group design features that increase the effectiveness of individual
and collaborative learning in student groups (Curşeu & Pluut, 2013).
One of the most extensively used design tool for groups is based on
the group role preferences described by Meredith Belbin (1981). Group
role preferences are defined as group members' predispositions to
these stable individual differences can configure in various ways within
groups (Belbin, 1981). The way in which individual role preferences com-
bine within groups is a configural group property (Klein & Kozlowski,
2000) that impacts on group dynamics and effectiveness. One of the
major claims of Belbin's role theory is that balanced groups (in which
all nine roles are present) perform better than unbalanced groups, in
which existing roles are duplicating each other (Belbin, 1981). Although
the claim of balanced groups is being extensively used in organizational
consultancy, the empirical evidence supporting its validity is not conclu-
sive (Blenkinsop & Maddison, 2007; Jackson, 2002; Partington & Harris,
1999; Senior, 1997; van de Water, Ahaus, & Rozier, 2008). The aim of
the current paper is to test the role balance claim in an educational setting
given that there is evidence that Belbin's group role theory can be applied
to non-managerial personnel as well (Fisher, Hunter, & Macrosson, 1998).
A comprehensive approach is being employed in which various group
balance indices are used to predict a wide array of outcomes in collabora-
tive learning groups: teamwork quality, group cognitive complexity, and
group performance, across time.

☆ The authors would like to thank Miel A. P. Vugts for helping with database manage-
ment and Marius T.H. Meeus for his useful comments on previous versions of the manu-
2. Theoretical underpinnings
script. Petru L. Curseu was supported by a grant of the Romanian National Authority for
Scientific Research, CNCS – UEFISCDI, project number PN-II-ID-PCE-2011-3-0482. The
funders had no role in study design, data collection and analysis, decision to publish, or Apart from their functional role (prescribed through design), group
preparation of the manuscript. members have the tendency to display particular behavioral patterns
⁎ Corresponding author at: Department of Organization Studies, Tilburg University, in interpersonal interactions that influence the progress of the group
Warandelaan 2, PO box 90153, 5000 LE Tilburg, The Netherlands. Tel.: +31 13466 8299.
E-mail addresses: m.n.meslec@uvt.nl (N. Meslec), petrucurseu@psychology.ro
towards specific task achievement. These stable individual differences
(P.L. Curşeu). are captured by the group role preferences (Belbin, 1981). The roots of
1
Tel.: +31 13 4662153; fax: +31 13 466 3002. group roles were considered to lie in a person's generic personality traits

http://dx.doi.org/10.1016/j.lindif.2015.03.020
1041-6080/© 2015 Published by Elsevier Inc.
82 N. Meslec, P.L. Curşeu / Learning and Individual Differences 39 (2015) 81–88
and mental abilities (Aritzeta, Swailes, & Senior, 2005; Belbin, 1981) as
well as the structure of environment (Arroba & Wedgwood-
Oppenheim, 1994; Fisher & Macrosson, 1995; Yuwei & Tang, 1997). In
the context of a 9-year research project developed by Belbin, behavioral
observations as well as personality and mental abilities of group members
were recorded and used to develop a taxonomy of group role preferences.
Matching of these measurements resulted in the identification of eight
group roles: the coordinator (CO — co-ordinates and controls the activi-
ties of the group), the resource-investigator (RI-extrovert, makes outside
contacts and develops ideas), the teamworker (TW-person oriented,
communicates well with the others), the plant (PL — creative and imagi-
native), the monitor-evaluator (ME — prudent and analytical), the imple-
menter (IM — practical and task-oriented), the completer–finisher (CF —
attentive to details, finishes things), and the shaper (SH — dynamic and
challenging). Later on, a ninth role was added to this taxonomy, the spe-
cialist (SP — with high technical skills and in-depth knowledge for the
task) (see Belbin, 1981, 2009 for an extensive overview).
One of the most important claims in Belbin's work is that balanced
groups (with regard to their members' role preferences) have superior
performance than unbalanced groups. In other words, it is useful to
have members that possess particular strengths without duplicating
the ones already present in the group (Belbin, 1981; van de Water,
van de Water, & Bukman, 2007). A perfectly balanced group would be
a group in which all nine roles are present in a high or very high level
while a perfectly unbalanced group would be one in which all the
group members report the same individual role preference. The concept
of role balance is therefore a configural property of groups (Klein &
Kozlowski, 2000), and the configuration of roles are expected to be pre-
dictive for group dynamics and performance (Belbin, 1981). According
to the Input–Process–Output model of group effectiveness (Ilgen,
Hollenbeck, Johnson, & Jundt, 2005), group role configurations can
be considered inputs that predict the group processes and the qual-
ity of interpersonal interactions, which in turn influence group per-
formance. However, the results of the studies investigating the impact of
group role balance on performance are not conclusive. While some
studies identified little or no relation between the two (Blenkinsop
& Maddison, 2007; Jackson, 2002; Partington & Harris, 1999; Senior,
1997; van de Water et al., 2008), some others found evidence for
group balance as a valid predictor of group performance (Higgss,
Plewnia, & Ploch, 2005; Prichard & Stanton, 1999).
The lack of converging results can be attributed to several factors.
First, in most of the studies there is no control for group size and gender
diversity although previous research reports gender differences regard-
ing group role preferences (Anderson & Sleap, 2004). Next to the gender
issue, some of the group balance indices are also sensitive to the group
size (van de Water et al., 2007). Therefore, gender and group size should
be accounted for when analyzing the relation between balance and
performance. Second, little specifications were given with respect to
how group role balance can be computed. Therefore, there is little overlap
in the formulas currently used in computing group role balance. Group
balance was computed by taking into account the behavioral or environ-
mental focus of the roles (Higgss et al., 2005), the weights of the top three
roles an individual in a group has (the so-called primary, secondary and
tertiary roles) (van de Water et al., 2008) or how much the group deviates
from an ideally balanced group (Partington & Harris, 1999). The different
assumptions underlying group balance formulas could stand as an expla-
nation for the non-conclusive results for the group balance-group perfor-
mance relation. Finally, most of the studies used just one group outcome
indicator, which often differed across studies. For instance, performance
was measured in terms of group processes such as group organization
and communication (Blenkinsop & Maddison, 2007; Prichard & Stanton,
1999) subjective reports of managers (Higgss et al., 2005) or group
success in simulation games (Partington & Harris, 1999; van de Water
et al., 2008).
In the current study the concept of group role balance and its impact
on performance is being reconsidered in order to tackle the limits
described above. Therefore, the impact of group role balance on group dy-
namics and outcomes is tested by using several balance indices (which
rely on different assumptions) as well as a variety of group dynamics
and performance indicators (group processes, group cognitive complexi-
ty, perceived and actual group performance).
3. The impact of group role balance on teamwork quality
Teamwork quality (TWQ) is a multidimensional construct that
reflects the quality of interpersonal interactions inside the group.
It consists of several dimensions that reflect both group processes
(communication, coordination and planning) and group emergent
states (cohesion, perceived performance and potency) (Curşeu & Pluut,
2013; Curşeu, Schalk, & Schruijer, 2010; Hoegl & Gemuenden, 2001).
Teamwork expertise and synergistic interactions within groups in
educational settings have several advantages. First, organizations are
often employing groups and teamwork as a flexible way of organizing
work and therefore are looking for candidates that already acquired
teamwork skills during their educational trajectory (Chen, Donahue, &
Klimoski, 2004). Thus, teamwork expertise increases workforce readi-
ness given that students develop during their studies specific teamwork
knowledge, skills and abilities. Second, when groups manage to develop
synergistic interactions (good teamwork quality) they generate com-
plex group-level knowledge structures (Curşeu & Pluut, 2013) as well
as increased group performance (in innovative projects) and at an indi-
vidual level are able to learn more (in terms of knowledge and skills)
and be more satisfied with their work (Hoegl & Gemuenden, 2001).
Given the benefits of teamwork quality, substantial effort was devoted
to investigate how the quality of interpersonal interactions within groups
can be improved in educational settings (Chen et al., 2004; Curşeu &
Pluut, 2013; Curşeu et al., 2012). Solutions such as the development of
university courses in which students specifically learn about teamwork
(e.g. The Psychology of Working in Groups and in Teams) (Chen et al.,
2004) or personality student-group interventions in which group mem-
bers learn about each other's personalities and how to manage individual
differences (Clinebell & Stecher, 2003) were proposed.
A more straightforward group design solution is the use of design
principles that generate the most effective group configurations.
However, the simple placement of students in groups does not al-
ways guarantee the development of teamwork skills (Hansen, 2006;
Johnson & Johnson, 1990). Student groups often experience unclear
goals, mismanagement, conflicts or unequal participation (Cox &
Bobrowski, 2000; McCorkle et al., 1999; McKendall, 2000; Rau &
Heyl, 1990). The aim of the current study is to investigate whether com-
posing groups by using the role balance assumption leads to better group
interactions and thus higher teamwork quality.
According to Belbin roles theory, group role balance, as a configural
group property should have a positive impact on teamwork quality. A cru-
cial role here is played by social roles such as coordinator and teamworker.
Due to their association with the extraversion dimension of personality
(Davies & Kanaki, 2006), such roles facilitate communication and coordi-
nation processes inside the group. The compromising style of conflict
management (Aritzeta et al., 2005) associated with these roles also buffers
emergent relationship conflicts and leads to positive emergent states such
as group cohesion. The resource-investigator, characterized by over-
optimism not only contributes to the group's belief in their own strengths
but, also helps in collecting and bridging among different ideas, including
those of introverted group members (e.g. plants or specialists).
Effective teamwork requires a balance between task related and in-
terpersonal knowledge as research on shared mental models argues
that members of effective groups need to share both task related as
well as teamwork related knowledge (Cannon‐Bowers & Salas, 2001).
The convergence of teamwork and taskwork mental models is condu-
cive to effective teamwork processes which in turn impact on group
performance (Mathieu, Heffner, Goodwin, Salas, & Cannon-Bowers,
2000). Moreover, meta-analytical evidence also suggests that shared
 N. Meslec, P.L. Curşeu / Learning and Individual Differences 39 (2015) 81–88
mental models have a positive influence on the quality of interpersonal
processes within groups (DeChurch & Mesmer-Magnus, 2010). A
balanced role composition secures that both task related (e.g.
specialist, plant) as well as teamwork related (e.g. teamworker, re-
source investigator) orientations are simultaneously present and
as such, balanced groups are expected to have more effective inter-
actions than unbalanced groups. Therefore, the fine adjustment of
one role to another in a balanced group should lead to synergistic
teamwork processes. We therefore hypothesize that:
H1. Group role balance positively predicts teamwork quality.
4. The impact of group role balance on group outcomes
Next to perceived group performance and objective performance
group cognitive complexity (GCC) is also used as an outcome variable.
Group cognitive complexity (GCC) was defined as ‘the richness of the
collective knowledge structures that emerge as a group-level phenom-
enon from the integration of individual specialized knowledge through
interpersonal interactions’ (Curşeu et al., 2010). The higher the cog-
nitive complexity, the stronger the group's capabilities to absorb the
variety of representations held by group members in relation to the
task or the environment (Neill & Rose, 2006) and the higher group per-
formance (Curşeu et al., 2010).
Research to date investigated how GCC and performance can be stim-
ulated in collaborative learning groups. Factors such as the composition
of groups with respect to gender, personal values, teamwork quality
and conflict were found to play a key-role for group's cognitive complex-
ity and performance (Curşeu & Pluut, 2013; Curşeu et al., 2012; Glew,
2009). In line with the claim developed by Belbin (1981), this study in-
vestigates whether role balance (as a configural group property) leads
to higher cognitive complexity and performance in groups. According
to Belbin (1981), it is beneficial to create groups in which individual char-
acteristics and role preferences are not redundant. A pure group (in
which all group members detain the same role) will develop a style of
its own, focused only on task or relational domains and such group role
configurations are not conducive to performance (e.g. a group of CW
will be well organized but with a lack of real ideas and inflexibility).
Therefore, a balanced group in which all nine roles are present will
contain all the characteristics needed for group performance. In line
with Belbin's balance claim, we hypothesize the following:
H2. Group role balance will positively predict group outcomes (GCC,
perceived and objective group performance).
5. Method
5.1. Sample and procedure
The sample consisted of 459 students (151 female) enrolled in an
Organizational Behavior course at a Dutch university. Participants
were asked to form groups at the onset of the course. Group size ranged
from three to seven members, resulting in 84 groups. The vast majority of
the groups had five or six members. For a detailed overview of demo-
graphics see Table 1. The more extreme group sizes (three and seven)
Table 1
Sample demographics.
Group size Number of women/group Total
0 1 2 3 4 5
3 1 0 0 1 0 0 2
4 0 4 1 0 0 0 5
5 4 8 15 7 0 0 34
6 4 6 17 6 3 1 37
7 1 1 2 1 0 0 5
83
resulted from the constraints imposed by the number of students enrolled
for each of the workshops. Groups had to work together throughout the
semester in order to deliver three group assignments that consisted of
three case studies that covered 40% of their individual grades. A cross-
lagged design was used, data being collected across a seven-week period
which was indicative for the duration of the course. Data collection in-
volved groups belonging to three different cohorts of students which
were taking the course in three distinct academic years. Group roles (as
a configural property of groups) were assessed in the first week of the se-
mester. Close to the end of the term (week 7), teamwork quality, per-
ceived performance and group cognitive complexity were assessed.
Objective performance was measured with three different group assign-
ments at three different time points (weeks 3, 5 and 7 of the course).
The separation of measurements in time has the advantage of reducing
the common method bias (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003).
5.2. Measures
5.2.1. Belbin roles questionnaire and role balance indices
Group roles were assessed using an adaptation of the role self-
assessment instrument developed by Belbin (BTRSPI) (Belbin, 1981).
The questionnaire (Belbin's team role questionnaire) was part of the
student's handbook and contained items referring to all nine Belbin
roles, including the specialist role which was not present in the 1981
version of BTRSPI (Huczynsky & Buchanan, 2001). The instrument has
seven sections and each section contains nine items, one for each of
the nine roles. For each section, participants had to distribute ten points
across the nine items, considering the behavior that best described him/
her in a group. A final score was computed for each group member for
each of the nine roles. Each role score was computed by adding up the
points allocated to the seven items that were referring to that particular
role. Theoretically, each role score can take values between 0 and 70.
In order to test the claim that balanced groups perform better, we
reviewed all the balance indices proposed in the literature and used
the ones that were presented detailed enough in order to be replicated.
In addition, a new balance index was added up and also other (thematic)
role patterns indices proposed in the literature. The balance indices (TB)
and role thematic patterns (TP) are described below.
The first balance index (TB1) relies on the assumption that in a bal-
anced group the aggregate score from all members will be evenly
spread across the nine roles (Partington & Harris, 1999). TB1 is comput-
ed with the following formula: TB1 = 9/[(∑ R − 7.77) + 9] ∗ 100,
where R represents the deviation of each of the nine roles, summed
up at a group level. Thus, TB1 considers the amount of roles directly at
a group level, disregarding their original configurations at an individual
level. A perfectly balanced group would have an R on each role equal to
7.77. This score represents the 70 points to be distributed divided by the
nine roles. If the total deviation of the role mean from 7.77 is 0, then
TB1 = 100%.
The premise for the second balance index (TB2 and TB2a) is that a
balanced group would have at least one person scoring high or very
high in as many as possible of the nine roles (Partington & Harris,
1999). For each role, norms for defining high and very high scores
were computed. TB2 uses the norms defined by Belbin (1981, p.158)
while TB2a uses the norms constructed by using the sample of this par-
ticular study. For each role, all the individual scores are ordered from the
smallest score to the highest. The scores are split in four quartiles, the
highest two quartiles (from 66 to 100%) containing the range of scores
which qualify as high and very high scores for that particular role. The
norms for this study were similar with the ones identified by Belbin
(1981, p.158) and Partington and Harris (1999, p. 700). Each group
received a maximum of one point for each role represented in a
high or very high level and two points if the role was not represent-
ed in a high or very high level. The following formula was used
TB2 = (9/∑ HR) ∗ 100%, where HR represents the sum of points
given for each role represented in a high or very high level.
84 N. Meslec, P.L. Curşeu / Learning and Individual Differences 39 (2015) 81–88
The third balance score (TB3) was computed on the premise that
ideally roles should not be duplicated (Belbin, 1981; Partington &
Harris, 1999). The formula used is similar with the one for TB2,
TB3 = (9/∑HR) ∗ 100%. The difference lies in the number of points at-
tributed to each role. In the case of TB3, each role represented in a high
or very high level counts for the number of points given.
Finally, the last balance index (TB4) considers group roles as a
qualitative property of the groups, in which both the richness of roles
represented in the group and the abundance of the roles making up the
richness are considered. Each group received one point for each role rep-
resented in a high or very high level and 0 points for roles not represented
as such. Group balance was computed using Simpson's index of diversity
1 - D = ∑n (n − 1)/N(N − 1), where n is the total number of high and
very high scores of a particular role and N is the total number of high and
very high scores of all roles (Simpson, 1949).
To summarize, four group balance indices were used, each of them
relying on different theoretical considerations. The first balance index
TB1 accounts for Belbin roles directly at a group level while TB2 and
TB2a consider roles at an individual level being afterwards configured
at a group level. TB3 is similar to TB2, bringing in addition a correction
for roles that duplicate each other. Finally, TB4 attempts to account for
Belbin roles in a more comprehensive way, while considering both the
variety of roles represented in the group as well as their elevation.
Next to the group role balance, Belbin classifies the proposed roles
in three thematic patterns (TP): relationship (CO, RI, TW), thinking
(PL, ME and SP) and action roles (IM, CF, SH) (Belbin, 2009). This last
distinction was also included in our analyses. For each thematic catego-
ry a group score was computed by averaging the individual scores to
the group level. For instance, for the relationship thematic pattern,
TPrelationship = (∑ COIndScores/GroupSize + ∑ RIIndScores/
GroupSize + ∑TWIndScores/GroupSize)/3, where IndScores represents
the individual score for that particular role.
5.2.2. Teamwork quality and group outcomes
5.2.2.1. Teamwork quality (TWQ). Groups had to rate their own
teamwork quality on a 5-point scale questionnaire in which they evaluat-
ed different dimensions of teamwork: communication, cohesion, coordi-
nation, planning and potency (Carron, Widmeyer, & Brawley, 1985;
Curşeu et al., 2010; Eby, Meade, Parisi, & Douthitt, 1999; Guzzo, Yost,
Campbell, & Shea, 1993). All items and their respective factor loadings
are presented in Table 2. Data was collected directly at a group level,
group members being asked to reflect on their own teamwork processes
during the semester and decide on a group rating for each of the
teamwork items. Data collection through group agreement was
argued to be superior to individual data collection aggregated sub-
sequently at a group level (Kirkman, Tesluk, & Rosen, 2001).The average
score of all the scales was used as an indicator for TWQ. An exploratory
factor analysis conducted with the scale items supports the one-
Table 2
Items and factor loadings for teamwork quality scale.
Items
“It was easy for all team members to express their ideas during team
discussions”
“Our group is united in trying to make high quality assignments”
“The group members feel they belong to this group”
“The group members share with each other the information they have
regarding the projects”
“The group members have synchronized their actions in order to reach
the group goals”
“Each group member was completely aware of his/her own tasks”
“The group has clearly defined performance standards”
“Each member is completely aware of the group's goals and objectives”
“This team has confidence in itself”
“This team believes it can be very productive”
dimensionality of the scale, with the first factor having an eigenvalue
of 3.33 and explaining for 33.30% of the total variance while the other
two factors identified explain only 14.13 and respectively 11.49 of the
total variance. This comes in line with previous studies indicating that
although teamwork quality has multiple facets its underlying structure
pertains to one unique factor (Eby et al., 1999; LePine, Piccolo, Jackson,
Mathieu, & Saul, 2008). The Alpha Cronbach for teamwork quality was
.75, which indicates a good reliability of the scale. The teamwork quality
scale was administered at the end of the course.
5.2.2.2. Group cognitive complexity (GCC). Group cognitive complexity
was measured through the cognitive map technique that is a group
task (Curşeu et al., 2010). Groups received 20 cards on which major con-
cepts from the course domain were written. Their task was to organize
the concepts in a way that reflects their understanding as a group of the
relations between the concepts. Once agreed on the final chart, they had
to glue the cards on a paper and draw the relations between the con-
cepts as well as to indicate the type of relations between the concepts.
The final maps were assessed while using three criteria: total number
of concepts used (NoC), total number of connections established
between the concepts (CMC) and the number of distinct relations
established (CMD). These three dimensions are indicative for group
cognitive complexity as the complexity of the conceptual systems in-
cluding on the one hand the number of dimensions belonging to the
system and one the other hand the nature and the extent of rules elicit-
ed for integrating these dimensions (Calori, Johnson, & Sarnin, 1994;
Curşeu et al., 2010). The type of relations was rated using seven different
categories: causal, association, equivalence, topological, structural,
chronological, and hierarchical (Curşeu et al., 2010; Gómez, Moreno,
Pazos, & Sierra-Alonso, 2000). Group cognitive complexity was comput-
ed using the following formula CMCo = (CMCxCMD)/NoC (Curşeu
et al., 2010).
5.2.2.3. Group performance. Perceived performance was measured with 2
items (Curşeu et al., 2010) on a 5-point Likert scale. One example of an
item is: “I think that the present group performance is”. For objective
performance the grades obtained by the groups at three different as-
signments were used. Groups had to deliver three assignments in
weeks 3, 5 and 7, which represented 40% of their final grade. In the
three assignments, groups had to analyze and solve three case studies
while using topics studied during the course (motivation at work, lead-
ership and group dynamics). The maximum grade for the first assign-
ment was 10 points while for the other two assignments they could
have earned a maximum of 15 points.
5.3. Control variables and data analysis strategies
In order to test the hypotheses several Ordinary Least Squares (OLS)
regression models were run with group balance indices and thematic
patterns as predictors and teamwork quality, group cognitive complex-
ity, perceived performance and objective performance (at three differ-
ent time points) as dependent variables. A separate OLS regression for
each of the dependent variables was run. Empirical evidence shows
Factor
that there are significant differences between men and women in
loadings
group roles assumed (Anderson & Sleap, 2004). In order to control for
.375 the potential influence of gender-based role preferences on our results,
.732
in the OLS we used the percentage of females in each of the groups as a
.463 control variable. Given that some group balance indices are sensitive to
.599 the number of members a group has (van de Water et al., 2007), group
size was used as a control variable.
.489
.414 6. Results
.490
.750 Perceived performance correlates positively with teamwork quality
.656 r(84) = .63, p = .00. and with objective performance at time 1, r
.662
(84) = .45, p = .00 (Tables 3 and 4). Group size is not associated with
 N. Meslec, P.L. Curşeu / Learning and Individual Differences 39 (2015) 81–88 85

Table 3
Bivariate correlations at group level.

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1. Group size
2. Percent females −.11
3. TB1 .06 .05
4. TB2 .46⁎⁎⁎ .00 .03
5. TB2a .26 ⁎⁎ .11 .06 .39⁎⁎⁎
6. TB3 −.32⁎⁎⁎ −.09 −.01 .13 .02
7. TB4 −.32⁎⁎⁎ .06 .08 .03 .53⁎⁎⁎ .33⁎⁎⁎
8. Relationship roles −.19 .00 −.36⁎⁎⁎ −.08 −.16 .19 −.21⁎⁎
9. Acting roles .09 .06 .19 .21 .05 −.16 −.05 −.47⁎⁎⁎
10. Thinking roles .07 −.07 .02 −.14 .09 −.02 .25⁎⁎ −.41⁎⁎⁎ −.59⁎⁎⁎
11. TWQ .11 .17 −.20 .02 −.10 .04 −.17 .22⁎⁎ .06 −.25⁎⁎
12. Perceived performance .23⁎⁎ .08 −.05 .09 .01 −.07 −.07 −.06 .18 −.11 .63⁎⁎⁎
13. GCC .15 .25⁎⁎ .28 ⁎⁎⁎ .16 .04 −.08 −.15 .04 .08 −.14 .14 .13
14. Performance 1 .22⁎⁎ .12 .28⁎⁎⁎ .10 .13 −.01 .06 −.24⁎⁎ .22⁎⁎ −.05 .18 .45⁎⁎⁎ .00
15. Performance 2 .03 .00 −.03 −.02 −.11 .00 −.02 −.04 −.06 .12 .12 .17 .09 −.19
16. Performance 3 .01 −.02 −.06 .03 .00 .02 .11 .00 .00 .02 .08 .12 .16 −.20 .61⁎⁎⁎
⁎⁎ p b .05.
⁎⁎⁎ p b .01.

TB1, r(84) = .06, p = .56, and positively associated with TB2, r(84) =
.46, p = .00, and TB2a, r(84) = .26, p = .00 while negatively associated
with TB3, r(84) = −.32, p = .00, and TB4, r(84) = −.32, p = .00. Per-
centage of women is positively associated to group cognitive complexi-
ty, r(84) = .25, p = .02. Finally, some significant correlations between
the group balance indices were identified: TB2 and TB2a, r(84) = .39,
p = .00, TB2a and TB4, r(84) = .53, p = .00, TB3 and TB4, r(84) = .33,
p = .00.
In order to test the proposed hypotheses a separate OLS regression
analysis was conducted for each of the dependent variables. Group
size and percentage of women were used as control variables in the
first step of the regression analysis. Team balance indices were added
in the second step of regression analysis. Due to multicolinearity issues,
the thematic patterns could not be included in one single OLS regres-
sion. Thus separate regression models were conducted for each of the
three thematic patterns.
The results indicate that TB1 positively predicts GCC (β = 0.28,
t(81) = 2.65, p = .01) and objective performance at time 1 (β = 0.24,
t(81) = 2.30, p = .02), and negatively predicts teamwork quality
(β = −.21, t(81) = −2.02, p = .04). All the other team balance indices
are not significantly related with any of the other team outcomes
(Tables 5 and 6). The three thematic configurations also do not predict
group perceived performance, objective performance or GCC. Neverthe-
less, thinking roles negatively predict teamwork quality (β = − .25,
t(81) = − 2.36, p = .02) while relationship roles positively predict
teamwork quality (β = .26, t(81) = 2.47, p = .01) (Tables 7 and 8). Fi-
nally, the percentage of women as a control variable is positively related
Table 4
Descriptive statistics.
to teamwork quality (β = .25, t(81) = 2.26, p = .02) and GCC (β = .24,
t(81) =2.24, p = .02).
7. Discussions
In the current study one of the major claims of Belbin's role theory
was tested, namely that balanced groups (in which all nine Belbin
roles are present) are associated with several group outcomes (group
cognitive complexity, perceived and objective performance) and team-
work quality. Among the team balance indicators, TB1 appears to be the
only to predict objective performance (measured at time 1) GCC and
teamwork quality (measured at the end of the group project). What dif-
ferentiates TB1 from the other balance indices is that roles are not con-
sidered at an individual level first and then summed up at a group level
but, rather as configurations of roles directly at a group level. For example,
for the coordinator role all the points belonging to all group members
(whether they are high or low) are being summed up and included in
the balance score. The rest of the formulas start from the assumption
that one particular role (e.g. coordinator) should be represented in a
high or very high level at one particular group member and that should
be considered for the group balance. The other low individual scores on
that particular role are not being accounted for.
TB1 was found to positively predict objective performance, but only
at time 1 and to be negatively related to teamwork quality as measured
at the end of the group project. We originally hypothesized that accord-
ing to Belbin's balance claim, the fine adjustment of one role to another
in a balanced group would lead to synergistic teamwork processes and
increased team performance. Furthermore, as research on shared men-
tal models argues, effective groups share in a balanced way both task as

Variable Mean SD Min Max

Table 5
1. Group size 5.46 .79 3 7 Impact of group balance on teamwork quality, perceived performance and cognitive
2. Percent females 33.66 19.88 0 100 complexity.
3. TB1 99.99 3.71 88.21 116.33
4. TB2 79.61 8.61 61.53 100 Step Variable Teamwork Perceived Cognitive
5. TB2a 81.87 6.37 66.67 88.89 quality performance complexity
6. TB3 54.76 7.38 38.10 72.73
β ΔR2 β ΔR2 β ΔR2
7. TB4 .90 .02 .85 .94
8. Relationship roles 8.03 .87 6.13 12.08 1 Group size .25 .02 .32⁎⁎ .04 .02 .06
9. Acting roles 8.81 .95 6.60 12.17 Percent females .25⁎⁎ .13 .20⁎
10. Thinking roles 6.47 .93 3.17 8.80 2 TB1 −.21⁎⁎ .07 −.08 0 .28⁎⁎ .11
11. TWQ 3.87 .41 2.80 4.80 TB2 −.06 −.01 .11
12. Perceived performance 3.86 .59 2.50 5.00 TB2a −.11 −.12 .07
13. GCC 2.64 .89 .60 5.70 TB3 .18 .01 .00
14. Performance 1 7.95 1.06 5 10 TB4 −.08 .08 −.22
15. Performance 2 12.19 2.18 5 15
⁎ p b .10.
16. Performance 3 11.79 1.77 7 15
⁎⁎ p b .05.
86 N. Meslec, P.L. Curşeu / Learning and Individual Differences 39 (2015) 81–88

Table 6 Table 8
Impact of group balance on objective performance at 3 different time points. Impact of role thematic configurations on objective performance at three different time
points.
Step Variable Performance 1 Performance 2 Performance 3
Step Variable Performance 1 Performance Performance
β ΔR2 β ΔR2 β ΔR2
2 3
1 Group size .33⁎⁎ .05 .18 0 .15 0
β ΔR2 β ΔR2 β ΔR2
Percent females .17 .05 −.00
2 TB1 .24⁎⁎ .08 −.04 0 −.09 0 1 Group size .22⁎⁎ .05 .04 0 .01 0
TB2 −.04 −.02 .02 Percent females .14 .01 −.01
TB2a −.05 −.24 −.18 2 Action roles .19⁎ .08 −.06 0 0 0
TB3 .06 .03 −.02 1 Group size .24⁎⁎ .05 .03 0 .02 0
TB4 .15 .16 .28 Percent females .17 .01 −.02
2 Thinking roles −.06 .04 .12 0 .02 0
⁎⁎ p b .05.
1 Group size .20⁎ .05 .02 0 .01 0
Percent females .15 .01 −.01
2 Relationship roles −.20⁎ .08 −.04 0 .01 0
⁎ p b .10.
well as teamwork related knowledge (Cannon‐Bowers & Salas, 2001). ⁎⁎ p b .05.
Our results indicate that although balanced teams perform significantly
better in initial phases of group development, this effect disappears to- accurately predict group objective performance or GCC. Nevertheless,
wards the end of the group project. This could be explained by other fac- thinking roles negatively predict teamwork quality while relationship
tors that have not been accounted for in the study. For example, the roles positively predict teamwork quality. Having teams composed most-
deterioration of interpersonal interactions in originally balanced groups ly of thinking oriented roles (plant, monitor evaluator and specialist)
could have buffered the positive relation between group role balance might interfere with the quality of interactions developed within groups.
and group performance at time points 2 and 3. The negative relation be- This comes in line with previous studies indicating that thinking roles are
tween TB1 and teamwork quality (measured at the end of the group characterized by a tendency towards domination in conflict, the use of
project) comes to support this claim. Studies indicate that initial group Machaivelian behaviors and fewer integrative cognitive styles (Aritzeta,
conflict is beneficial for group performance as it stimulates information Swailes, & Senior, 2007; Aritzeta et al., 2005; Macrosson & Hemphill,
exchange but when it escalate in more profound relationship conflicts it 2001). On the other hand relationship oriented roles (coordinator, re-
becomes detrimental for the group (Yang & Mossholder, 2004). It could source investigator and teamworker) might facilitate the quality of inter-
be that balanced groups, engaged in constructive task conflicts at the actions within groups. Relationship roles were shown to correlate with
onset of the group project, yet in time, task conflict evolved into rela- the extraversion dimension of personality (Davies & Kanaki, 2006) and
tionship frictions that damaged the quality of interpersonal interactions bridging and adaptive cognitive styles (Aritzeta et al., 2007; Fisher et al.,
and ultimately group performance. Thus, the disappearing effect of 1998). Further on, group members adopting relationship roles are more
team balance on group performance from time 1 to time 2 and 3 could likely to adopt avoiding and compromising conflict management styles
be due to process losses experienced by the group members. Further (Aritzeta et al., 2005). These characteristics of relationship roles might fa-
studies could investigate the mechanisms at play here. Interestingly, al- cilitate the quality of interpersonal interactions within groups.
though balanced groups do not manage to maintain the same levels of Percentage of women was found to be significantly related to team-
performance at later stages, they do develop a higher level of cognitive work quality and GCC. One possible explanation for this finding can be
complexity than unbalanced groups. TB1 positively predicts the level of that women display a stronger preference for relationship roles than
GCC (as measured at the end of the group project). GCC was shown pre- men. This claim is in line with previous empirical evidence that
viously to be a stable group emergent state and for this reason might be women have a stronger preference for the teamworker role than men
more resilient to changes or negative interaction patterns (Curşeu et al., (Anderson & Sleap, 2004; Sommerville & Dalziel, 1998). The same pref-
2010). Although balanced groups might be able to consider a higher num- erence was also found in our study t = 2.09, p b .05 (mean scores for
ber of relations between relevant knowledge structures and thus to have women = 9.24 and mean scores for men = 8.24). An alternative expla-
a richer representation of the task they might still have difficulties in nation can be linked to the development of collective emotional intelli-
translating this into performance due to negative interactions developed gence in groups. Recent studies indicate that the percentage of women
at later stages of group development. in the group fosters the emergence of collective emotional intelligence
Next to team balance indices, the impact of the three thematic config- that further increases social integration (high cohesion and reduced
urations (action, relationship and task related roles) on group outcomes relationship conflict) within groups (Curşeu, Pluut, Boroş, & Meslec, in
was also tested. Our results indicate that thematic configurations do not press).
Finally, a few correlational findings warrant attention. Perceived
Table 7 performance was found to significantly correlate with both teamwork
Impact of role thematic configurations on teamwork quality, perceived performance and quality and objective performance at time 1. This comes in line with
cognitive complexity.
previous studies indicating that teamwork processes are positively as-
Step Variable Teamwork Perceived Cognitive sociated with both group performance and members' satisfaction
quality performance complexity (LePine et al., 2008). Group size is not associated with TB1 and positively
β ΔR2 β ΔR2 β ΔR2 associated with TB2 and TB2a. This indicates that larger groups have a
higher chance of being role-balanced, which involves the representa-
1 Group size .13 .02 .22⁎⁎ .04 .14 .06
Percent females .18⁎ .10 .24⁎⁎ tion of all nine roles at a high or very high level. However, this reasoning
2 Action roles .03 .01 .15 .05 .04 .05 does not hold for TB1, an index in which roles are counted for directly at
1 Group size .15 .02 .25⁎⁎ .04 .18⁎ .06 a group level. Thus, group size is not as relevant for TB1 as it is for TB2
Percent females .17 .10 .23⁎⁎
and TB2a. Further on, group size is negatively correlated to TB3 and
2 Thinking roles −.25⁎⁎ .07 −.13 .04 −.13 .07
1 Group size .18⁎ .02 .24⁎⁎ .04 .16 .06 TB4. TB3 and TB4 although similar to TB2 in counting the roles which
Percent females .19⁎ .11 .25⁎⁎ are represented at a high or very high level, also include corrections
2 Relationship roles .26⁎⁎ .08 −.01 .03 .07 .05 for roles that are duplicating each other and respectively for the intensi-
⁎ p b .10. ty of roles. These corrections could explain why the link with group size
⁎⁎ p b .05. is negative. TB1 does not correlate with any of the other balance indices,
 N. Meslec, P.L. Curşeu / Learning and Individual Differences 39 (2015) 81–88
given that the index is constructed by using a different assumption,
namely that group roles are considered directly at a group level. Per-
centage of women positively correlates with group cognitive complexi-
ty. This comes in line with previous findings indicating the important
role of gender for groups and group performance (Curşeu et al., in
press; Woolley, Chabris, Pentland, Hashmi, & Malone, 2010).
8. Implications, limitations and directions for further research
Our study has important implications for group design literature as
well as for collaborative learning literature. First, we report the results
of a comprehensive test of various group configurations with respect
to group role preferences on various collaborative learning outcomes.
We find that team balance is relevant at least for initial levels of group
performance when it is considered as a group configural property rather
than a sum of individual roles. Further on, balanced groups are associated
with higher levels of cognitive complexity. This indicates that forming
groups based on Belbin role balance assumption might enhance group
performance, at least in initial phases and have persistent effects on the
level of group cognitive complexity. Further research could explore here
other intervening processes such as group conflict that might explain
the drop in performance from time 1 to time 2 and 3. Additionally, we
find that the percentage of women positively predicts teamwork quality
and GCC. We admit that using the percentage of women as a design prin-
ciple when forming student groups in educational settings has limitations,
yet a straightforward implication of our results is that educators should at
least attempt to form mixed gender groups. This is in line with previous
results showing that gender composition has positive influence on GCC
(Curşeu, Schruijer, & Boroş, 2007) and group rationality (Curşeu, Jansen
& Chappin, 2013). When gender composition is not a feature open to ma-
nipulation, educators could focus on short trainings to increase interper-
sonal awareness and social sensitivity or use normative interventions to
influence directly the quality and nature of interpersonal interactions in
collaborative learning groups (Curşeu & Schruijer, 2012).
Future research could explore possible explanatory mechanisms and
our study points out that the quality of interpersonal interactions needs
further attention in this respect. Finally, our study uses data collected
from multiple sources (e.g., self-rated performance, objective perfor-
mance indicators and group cognitive complexity rated by external
evaluators) and measures are separated in time, therefore the likely im-
pact of common method bias is low. Given the strengths of our design,
the substantial sample size and the variety of outcomes considered we
can conclude that our findings are robust, yet future research is wel-
comed in replicating them in other organizational contexts.
Next to these strengths, our study has also certain limitations. First,
group roles were measured at the beginning of the workgroup as indi-
vidual preferences for particular behavioral styles. However, we did
not check whether these were indeed the roles assumed by the group
members during their work. Although group roles are in principle con-
sidered to be individual preferences that are stable over time, it could be
the case that due to dynamics unfolded in the group, members take
other roles (e.g. secondary or tertiary roles) than the ones they have
their first preference for. Second, the self-reported nature of the team-
work measure might be a limitation of the study. Nevertheless, in
using this scale we do align with previously proposed measures for
teamwork quality (Eby et al., 1999; LePine et al., 2008). Third, our
groups differed in size and this might be problematic especially for the
computation of group role balance that involves the presence in a high
or very high level the presence of all nine roles. Although we tried to
overcome this limit by controlling for group size in our analyses and
by using different group balance indices, we do acknowledge this as a
limitation of our study. Further studies could try to replicate these
results in a more controlled setting in which all groups are composed
with the same size and group roles preference stability is accounted
for. Finally, in this study group role balance was not manipulated and
thus causal inferences are not warranted. Further studies could try to
87
explicitly compose groups based on role preferences and investigate
the impact of group role balance on group outcomes.
9. Conclusions
Belbin group roles are widely used in organizations, although the
initially proposed claims have created many scientific controversies
(Belbin, 1993; Furnham, Steele, & Pendleton, 1993). Our study indicates
that when group role balance is conceptualized as a configural group
property instead of a sum of individual roles, it positively predicts
group performance at least at initial phases of the group work and it
positively predict group cognitive complexity. Nevertheless it links neg-
atively with teamwork quality and it does not predict performance at
later points in time. Further studies could investigate the dynamic ef-
fects of role balance over time on group performance. Next to this,
other group diversity indicators such as the percentage of females in
the group appear to be highly predictive for teamwork quality and
group cognitive complexity and thus should warrant further research.
References
Anderson, N., & Sleap, S. (2004). An evaluation of gender differences on the Belbin team
role self-perception inventory. Journal of Occupational and Organizational Psychology,
77, 429–437.
Aritzeta, A., Swailes, S., & Senior, B. (2005). Team roles: Psychometric evidence, construct
validity and team building. Research Memorandum.
Aritzeta, A., Swailes, S., & Senior, B. (2007). Belbin's team role model: Development, valid-
ity and applications for team building. Journal of Management Studies, 44(1), 96–118.
Arroba, T., & Wedgwood-Oppenheim, F. (1994). Do senior managers differ in the public
and private sectors? Journal of Managerial Psychology, 9(1), 13–16.
Belbin, M. (1981). Management teams, why they succeed or fail. London: Heinemann.
Belbin, R.M. (1993). A reply to Belbin team role self-perception inventory by Furnham,
Steele and Pendleton. Journal of Occupational and Organizational Psychology, 66,
259–260.
Belbin, M. (2009). Method, reliability & validity, statistics and research: A comprehensive
review of Belbin team roles. Retrieved from www. belbin.com
Blenkinsop, N., & Maddison, A. (2007). Team roles and team performance in defence ac-
quisition. Journal of Management Development, 26(7), 667–682.
Burch, G.S.J., & Anderson, N. (2004). Measuring person–team fit: Development and
validation of the team selection inventory. Journal of Managerial Psychology, 19,
406–426.
Calori, R., Johnson, G., & Sarnin, P. (1994). CEOs' cognitive maps and the scope of the
organization. Strategic Management Journal, 15, 437–457.
Cannon‐Bowers, J.A., & Salas, E. (2001). Reflections on shared cognition. Journal of
Organizational Behavior, 22(2), 195–202.
Carron, A.V., Widmeyer, W.N., & Brawley, L.R. (1985). The development of an instrument
to assess cohesion in sport teams: The Group Environment Questionnaire. Journal of
Sport Psychology, 7, 244–266.
Chen, G., Donahue, L.M., & Klimoski, R.J. (2004). Training undergraduates to work in orga-
nizational teams. Academy of Management Learning and Education, 3, 27–40.
Clinebell, S., & Stecher, M. (2003). Teaching teams to be teams: An exercise using the
Myers-Briggs® type indicator and the five-factor personality traits. Journal of
Management Education, 27(3), 362–383.
Cox, P.L., & Bobrowski, P.E. (2000). The team chapter assignment: Improving the effec-
tiveness of classroom teams. Journal of Behavioral and Applied Management, 1,
92–103.
Curşeu, P.L. (2011). Need for cognition and active information search in small student
groups. Learning and Individual Differences, 21(4), 415–418.
Curşeu, P. L., Jansen, R. J., & Chappin, M. M. (2013). Decision rules and group rationality:
cognitive gain or standstill? PloS one, 8(2), e56454.
Curşeu, P.L., Janssen, S.E.A., & Raab, J. (2012). Connecting the dots: Social network struc-
ture, conflict, and group cognitive complexity. Higher Education, 63(5), 621–629.
Curşeu, P.L., & Pluut, H. (2013). Student groups as learning entities: The effect of group di-
versity and teamwork quality on groups' cognitive complexity. Studies in Higher
Education, 38(1), 87–103.
Curşeu, P.L., Pluut, H., Boroş, S., & Meslec, N. (2015). The magic of collective emotional
intelligence in learning groups: No guys needed for the spell! British Journal of
Psychology (in press).
Curşeu, P.L., Schalk, R., & Schruijer, S. (2010). The use of cognitive mapping in eliciting and
evaluating group cognitions. Journal of Applied Social Psychology, 40(5), 1258–1291.
Curşeu, P., & Schruijer, S. (2012). Normative interventions, emergent cognition and deci-
sion rationality in ad-hoc and established groups. Management Decision, 50(6),
1062–1075.
Curşeu, P.L., Schruijer, S.G.L., & Boroş, S. (2007). The effects of groups' variety and disparity
on groups' cognitive complexity. Group Dynamics: Theory, Research and Practice,
11(3), 187–206.
Davies, M., & Kanaki, E. (2006). Interpersonal characteristics associated with different
team roles in work groups. Journal of Managerial Psychology, 21(7), 638–650.
DeChurch, L.A., & Mesmer-Magnus, J.R. (2010). The cognitive underpinnings of effective
teamwork: A meta-analysis. Journal of Applied Psychology, 95(1), 32.
88 N. Meslec, P.L. Curşeu / Learning and Individual Differences 39 (2015) 81–88
Eby, L.T., Meade, A.W., Parisi, A.G., & Douthitt, S.S. (1999). The development of an individ-
ual level teamwork expectations measure and the application of a within group
agreement statistic to assess shared expectations for teamwork. Organizational
Research Methods, 2, 366–394.
Fisher, S.G., Hunter, T.A., & Macrosson, W.D.K. (1998). The structure of Belbin's team roles.
Journal of Organizational and Occupational Psychology, 71, 283–288.
Fisher, S.G., & Macrosson, W.D.K. (1995). Early influences on management team roles.
Journal of Managerial Psychology, 10(7), 8–15.
Furnham, A., Steele, H., & Pendleton, D. (1993). A response to Dr. Belbin's reply. Journal of
Organizational and Occupational Psychology, 66, 261.
Glew, D.J. (2009). Personal values and performance in teams: An individual and team-
level analysis. Small Group Research, 40(6), 670–693.
Gómez, A., Moreno, A., Pazos, J., & Sierra-Alonso, A. (2000). Knowledge maps: An essential
technique for conceptualisation. Data and Knowledge Engineering, 33, 169–190.
Guzzo, R.A., Yost, P.R., Campbell, R.J., & Shea, G.P. (1993). Potency in groups: Articulating a
construct. British Journal of Social Psychology, 32, 87–106.
Hansen, R.S. (2006). Benefits and problems with student teams: Suggestions for improv-
ing team projects. Journal of Education for Business, 82(1), 11–19.
Haugwitz, M., Nesbit, J.C., & Sandmann, A. (2010). Cognitive ability and the instructional
efficacy of collaborative concept mapping. Learning and Individual Differences, 20(5),
536–543.
Higgss, M., Plewnia, U., & Ploch, J. (2005). Influence of team composition and task com-
plexity on team performance. Team Performance Management, 11, 227–250.
Hoegl, M., & Gemuenden, H.G. (2001). Teamwork quality and the success of innovative
projects: A theoretical concept and empirical evidence. Organization Science, 12(4),
435–449.
Huczynsky, A., & Buchanan, D. (2001). Organizational behaviour: An introductory text (4th
ed.). Essex: Pearson Education Limited.
Ilgen, D.R., Hollenbeck, J.R., Johnson, M., & Jundt, D. (2005). Teams in organizations: From
input–process–output models to IMOI models. Annual Review of Psychology, 56,
517–543.
Jackson, C. (2002). Predicting team performance from a learning process model. Journal of
Managerial Psychology, 17(1), 6–13.
Johnson, D.W., & Johnson, R.T. (1990). Cooperative learning and achievement. In S. Sharan
(Ed.), Cooperative Learning: Theory And Research (pp. 23–37). New York: Praeger
Press.
Kirkman, B.L., Tesluk, P.E., & Rosen, B. (2001). Assessing the incremental validity of team
consensus rating over aggregation of individual-level data in predicting team effec-
tiveness. Personnel Psychology, 54, 645–667.
Klein, K.J., & Kozlowski, W.J. (2000). From micro to meso: Critical steps in conceptualizing
and conducting multilevel research. Organizational Research Methods, 3(3), 211–236.
LePine, J.A., Piccolo, R.F., Jackson, C.L., Mathieu, J.E., & Saul, J.R. (2008). A meta‐analysis of
teamwork processes: Tests of a multidimensional model and relationships with team
Effectiveness criteria. Personnel Psychology, 61(2), 273–307.
Macrosson, W.D., & Hemphill, D.J. (2001). Machiavellianism in Belbin team roles. Journal
of Managerial Psychology, 16(5), 355–363.
Mathieu, J.E., Heffner, T.S., Goodwin, G.F., Salas, E., & Cannon-Bowers, J.A. (2000). The in-
fluence of shared mental models on team process and performance. Journal of Applied
Psychology, 85(2), 273.
McCorkle, D.E., Reardon, J., Alexander, J.F., Kling, N.D., Harris, R.C., & Iyer, R.V. (1999). Un-
dergraduate marketing students, group projects, and teamwork: The good, the bad,
and the ugly? Journal of Marketing Education, 21, 106–117.
McCune, V., & Entwistle, N. (2011). Cultivating the disposition to understand in 21st
century university education. Learning and Individual Differences, 21(3), 303–310.
McKendall, M. (2000). Teaching groups to become teams. Journal of Education for Business,
75, 277–282.
Neill, S., & Rose, G.M. (2006). The effect of strategic complexity on marketing strategy and
organizational performance. Journal of Business Research, 59, 1–10.
O'Neil, H.F., Allerd, K., & Baker, E.L. (1997). Review of workforce readiness theoretical
frameworks. In H.F. O'Neil (Ed.), Workforce readiness: Competencies and assessment
(pp. 3–25). Mahwah, NJ: Erlbaum.
Partington, D., & Harris, H. (1999). Team role balance and team performance: An empir-
ical study. Journal of Management Development, 18(8), 694–701.
Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., & Podsakoff, N.P. (2003). Common method
biases in behavioral research: A critical review of the literature and recommended
remedies. Journal of Applied Psychology, 88(5), 879.
Prichard, J.S., & Stanton, N.A. (1999). Testing Belbin's team role theory of effective groups.
The Journal of Management Development, 18(8), 652–665.
Rau, W., & Heyl, B.S. (1990). Humanizing the college classroom: Collaborative learning
and social organization among students. Teaching Sociology, 18, 141–155.
Senior, B. (1997). Team roles and team performance: Is there ‘really’ a link? Journal of
Occupational and Organizational Psychology, 70, 241–258.
Simpson, E.H. (1949). Measurement of diversity. Nature, 163, 688.
Sommerville, J., & Dalziel, S. (1998). Project teambuilding—The applicability of Belbin's
team-role self-perception inventory. International Journal of Project Management,
16(3), 165–171.
Stevens, M.A., & Campion, M.J. (1999). Staffing work teams: Development and validation
of a selection test for teamwork settings. Journal of Management, 25, 207–228.
Tannenbaum, S.I., Mathieu, J.E., Salas, E., & Cohen, D. (2012). Teams are changing: Are
research and practice evolving fast enough? Industrial and Organizational
Psychology, 5, 2–24.
van de Water, H., Ahaus, K., & Rozier, R. (2008). Team roles, team balance and perfor-
mance. Journal of Management Development, 27(5), 499–512.
van de Water, T., van de Water, H., & Bukman, C. (2007). A balanced team generating
model. European Journal of Operational Research, 180, 885–906.
Woolley, A.W., Chabris, C.F., Pentland, A., Hashmi, N., & Malone, T.W. (2010). Evidence for
a collective intelligence factor in the performance of human groups. Science,
330(6004), 686–688.
Yang, J., & Mossholder, K.W. (2004). Decoupling task and relationship conflict: The role of
intragroup emotional processing. Journal of Organizational Behavior, 25(5), 589–605.
Yuwei, S., & Tang, H.K. (1997). Team role behaviour and task environment: An explorato-
ry study of five organizations and their managers. Journal of Managerial Psychology,
12(2), 85–94.
Zedeck, S., & Goldstein, I.L. (2000). The relationship between I/O psychology and public
policy: A commentary. In J.F. Kehoe (Ed.), Managing selection in changing organizations:
Human resource strategies (pp. 371–396). San Francisco, CA: Jossey- Bass.